Welcome to NexusFi: the best trading community on the planet, with over 150,000 members Sign Up Now for Free
Genuine reviews from real traders, not fake reviews from stealth vendors
Quality education from leading professional traders
We are a friendly, helpful, and positive community
We do not tolerate rude behavior, trolling, or vendors advertising in posts
We are here to help, just let us know what you need
You'll need to register in order to view the content of the threads and start contributing to our community. It's free for basic access, or support us by becoming an Elite Member -- see if you qualify for a discount below.
-- Big Mike, Site Administrator
(If you already have an account, login at the top of the page)
Before I say anything, I would like to give a huge thanks to the following members for providing such wonderful indicators and advice. @dorschden, @Fat Tails, @Big Mike, @vvhg, @Zondor, @NJAMC, @dorschden [again, you get 2] , @TheWizard, @gomi, @Xav1029 ,(even though you're a Gator fan), and the person I owe most of my trading knowledge to, @Kazonomics. Without you buddy, i'd still be manually losing my ass off. And there are others but it's mostly your code that inspired me to do some really interesting strategic programming. And of course, the folks at TopStep @TopstepTrader for this wonderful program.
I've been an unsuccessful trader for about 10 years now, but a quite successful software engineer for 15. So a few years ago I started working on automated systems. I've used a variety of SDK's but settled on NinjaTrader as I'm able to develop everything I need in Visual Studio using the NT functions and the entire windows platform for an OOP approach to development. For data storage I started with SQL Server, but this was just too slow to run in real time, so I switched to Mongo. I'm sure I could have use MySql or better code optimization (most likely), but Mongo serves it's purpose well. Not gonna do a whole bio here but I've done quite a bit in my 38 years of existence. From producing music, owning a record label, theoretical research on classical and quantum mechanics, Dj'ing at nightclubs all over the world, and working at Burger King(first job at 15). That's me - sorta.
Now for today - It was a day of leaving a LOT of money on the table, but I ended up $765 so I' not going to complain. But As you'll see in the charts below it could have been a lot better. The systems I've programmed are fully automated with about 5 different rule sets integrated for different signals. I run these against the 6A, 6B, 6C, 6E, 6N, 6S, YM, NQ, ES, GC, & the CL. I run the indexes from 8:15Am to 10:30Am and Currencies, GC and CL from 3:30 AM to 11:55 AM. This is after exhaustive back testing to find the best time for each signal. There are a few signals I only run on certain instruments. I'll get into each system in a later post. But they are all based on Price Action, breakouts, mean reversion, arbitrage, extensive statistical analysis and a little bit of ML i've designed using Encog - an SVM Classification system that i've trained to predict whither the next swing will reach my targets - this is used on one signal only.
After having a 48 hour or so coding marathon to finish things up for today, I fell asleep in my chair and never enabled the strategy that listens for signals sent from my charts to place trades. So I missed a really good morning. But ~10 when I check to see how i was doing in the overnight, I realized i was an idiot, turned things on and nailed a few good ones.
At the moment my exit strategies suck. I'm using NT's ATM's. But I'm in the process of developing exit strategies that will realize a lot more profit on the good trades, kick me out of the losers early, along with a comprehensive risk management system. For the exit strategies I'm using the wonderful Local Order Manager . @NJAMC you rock! I'll post my finished product for all to use.
Trade one on the NQ was a bust. I can't seem to show the entries and exits on the chart, but I'm posting the trade list from NT for confirmation.
Here are the rest of the signals. All realtime signals are marked with an arrow as they happen. Each color is a different strategy. I run all the strategies on different Renko tic sizes (I lazily don't do this in code)
YM 9 and 10 Tick Renko
6S 4-5-6 Tick
GC 10 Tick - MISSED - Pointing to wrong account
6N - MISSED, still trying to figure out why I didn't trigger.
And here are the final day's stats. Not bad, but like I said - once I get the LOM exit strategy completed and my technical sleepiness worked out, it should be better going forward.
Thanks for reading and I'll keep these going every day. I'll get into the strategies and code as the days go on. have a awesome evening and may the force be with you. - Mike
Yes you are correct & its the 10 day). It's been 2 days and I'm up ~1450. Gonna post Fridays results today. Next week should be better. I'll make it happen. Finishing up the trade management portion of the software this weekend (exit strategies mostly) and that should help realize a lot more gains. That damn NQ. Only 2 losses are from that. Wrote a new mtf filter (thanks to some @Fat Tails code!) last night that should keep me out of triggering buys at small pullbackss at the top of swings.
I'm keeping consistent. Short post today - got a busy one ahead of me. Nailed 6 out of 7. Looking forward to seeing how things perform on downtrend days though. thurs and fri have been pretty good. Below are the results.
Loser NQ again!
GC
CL - Missed the first one(looking into why - think I made a code change in VS and that conflicted with NT) - hit the second. and left a lot on the table. No more after this weekend though. Finishing up my exit strat. Also, will explain the red/white bars next week. Wrote a statistical trend filter (mostly for CL) based an about 30 parameters weighted on performance over f(x) time period to score the bar. score1>score2 = red and vice-versa. MTF swing size, duration, length, relation, divergence and ~25 others things.. Side note: these are also the params i use to train my SVM. When I clean and test the the code a bit more I'll upload it.
Hello trader friends! Sorry I haven't posted all week. It's been a long one. Between managing this, my software work (full time contract), driving to my parents house last weekend for a visit, finishing up this gd order management system (ugh - will get into that), attempting to maintain what social life I have left, and just work work work. Notice sleep is not in there.
While everyone was drinking beer and grilling out at my Dad and Stepmom's new place last weekend, I'm at the kitchen table hammering a way at code. And drinking beer. I had to get it done.
The Order Management System: Time to try this sucker out in realtime - a Strategy that acts as a "listener" tied to the charts where the entry logic is. I couldn't use the LOM for the Combine as they require a market stop at all times, so an unmanaged approach wasn't feasible. Although, I will use it for my $$ trading and my trading system for Top Step if I'm a chosen one. So I built it using the old IOrder and all the other interfaces that work with managed orders. The idea was simple.
1. Listen for trade signals running in the indicator strategy.
2. Based on the instrument, set a stop for all orders and profit target for all but one. The third would be a runner.
3. I didn't want to use a trail so I built a FAST Moving Average that I would use as an exit signal.
#3 Simple premise:
After 2 bars since entry start reading the ma output. If the Slope(DS,2,0)<0 get out! Got all the code working and ran it against market replay for the previous week. WOW, I could have tripled my gains. Exciting! So i put it into action on monday. That's when i realized it didn't work like expected. Got kicked out of trades early, and sometimes the MA lost the handle of the contract it was watching. Well Lange, that's what you get for writing a single system to manage 9 instruments with 5 potential entries on 22 different charts - while coding in record time. Not spaghetti but angel hair in this one!. Note to self: Re-write the strategies using same logic, but in separate indicators. Well, I have too. Managing 9507 lines of code across 5 strategies in a single file is not best practices i'm assuming. And that's not counting the external indicator code (most of which i owe thanks to the wonderful members here).
I used my new OM system Monday and Tuesday then reverted back to ATM's(ugh!) for Wed, Thursday and today. I'll debug(re-write) the hell out of it this weekend and see what happens next week. The only charts I'm going to show are todays. but I have the NT trade results to show for all the days.
"Am I going MAD, or did I hear the word THINK escape your lips?" -vizzini
MONDAY:
Got three signals, 2 on the YM and one on the ES for a total of 9 contracts. And good signals they were. But my good old order management system kicked me out way too early. Broke me even on the ES - frustrating. But it was a + $345 day at 100%. So i don't feel too bad. Can't. Take a win as a win.
TUESDAY:
5 Signals and all profitable! I'm on a roll. Up 620 and feeling good. Then the 10ish YM signal came in. The MA didn't pick it up and ended up loosing 500 on that single trade. Down to 120 - paid for my commissions. But still an up day. Not too happy about it. But not a loser.
WEDNESDAY:
Big loser. I'm not making excuses, this happens, but I did what you're NEVER supposed to do with ANYTHING! If it's not broken, don't fu&#*ng fix it. Bored tuesday night and decided, I'm not making enough here to win this Combine. Let's manipulate some code! Well, the images below show the results of that debacle. after 3 losses and -1000 I shut down. Computerly and emotionally. But suck it up Lange, learn from that, revert to working code and move on.
THURSDAY:
10 Signals and 40 contracts. And got my arse whipped by the 6B. But ended up $525. Always been a fan of the English, some of whom have been and still are my best friends in the world. But today I wanted to have tea with the Queen - hide the earl grey or something.
FRIDAY!:
Good day! Other than getting slapped by the 6B AGAIN, ended up ~1750. Although I did break one of my rules. Never click trade! But I had to short the GC. Had to. Too perfect. And I scalped a little on the 6J and 6B - STOP PRESSING BUTTONS!! The ES should have signaled too, but it didn't. Gotta find out why.
GC - My rule break that paid off.
Nice little move on the CL - And you can see yesterday's on this chart too
My nemesis the 6B
New signal Big Red comes though on the NQ (slight derivative of Hot Pink)
The YM and 4 trades . 5 really, but I got out early on Big Red later in the day on the 9Tick. Golden Boy gave me three signals. 2 on the the 9Tick YM (two in a row says, "this is a winner buddy") and one on the 10Tick. And old faithful Hot Pink gave me one on the 10tick.
YM 9Tick
YM 10Tick
Totals for today
So far - not a bad 7 days. The combine rules are strict, so unless I can grab $700 on the 6B (to keep my positive per instrument) and about 8K more in profits, it doesn't look like I'll be a winner. But I have 3 days left. and hell, If anyone else can do it, so can I. Regardless of what happens, I'm damn proud of myself and (finally) built what seems to be profitable automated trading strategies (after 10 years of losing - but learning!). yea, only 7 days and they are Long only strategies(for now).
Till monday! have a great weekend everyone!!
Mike
"Pick a flower on Earth and you move the farthest star. " -Paul Dirac
